Serverless, of course, does not imply that there is no server. I prefer more the term Function-as-a-Service, but FaaS is still one method of serverless computing after all. https://martinfowler.com/articles/serverless.html The blog post is more about experimenting, and knowing how things work under the hood, and that you don't need AWS Lambda for example if you want to work with serverless, but you can do it at home with your RPi.

First you are told "serverless" doesn't mean there is no server, because in fact you are using servers, it's just that administering those servers is somebody else's job.

Now apparently even setting up a server yourself can be "serverless". It goes to show that terms like these mean "blah-blah-blah" to people in general.

See also the discussion about "isomorphic".

Yup, but with extra steps to make it "serverless". It's cool, but I don't see the point.

It's more about experimenting, and knowing how things work under the hood. So the learning part is the poing. But also that you can have a lightweight serverless platform, that doesn't require a lot of resources since OpenFaas requires Kubernetes or Docker Swarm, but with faasd you can have it on a single RPi.

If you write a serverless server correctly then when you run it, the server disappears but the code keeps running.